ok, it appears I have the threading issues better so we can call force refresh after changing clusters

    processes (e.g. "/proc/bus").

the patch is getting too big, things still work for sview.  Quite a bit of infrastructure change to allow sview to work cross-cluster

    as needed based upon the processs's resource limits.

we still need work on the step retry when gres resources are all busy

a problem still remains with respect to oversubscribing a job's gres allocation at the step level